Diagnosing a problem on a 7037-A50 or a 7047-185

This is the entry procedure for diagnosing problems on a 7037-A50 or a 7047-185

If possible, leave the server in the state that it was in when the problem first occurred. Before you begin, visually check the system for symptoms, damage, or obvious problems such as unplugged or damaged power cables, correct external device cabling, or external devices that are powered off. Correct any problems that you find. Make a note of any symptoms that your server displays, such as:

the system-attention light emitting diode (LED) is continuously illuminated.
a reference code is displayed in the control (operator) panel display.
you are having a problem with an adapter or an attached device.

Note: Because your server stores information about system operations even after the system unit is powered off, do not unplug the power cord from the system unit (or the power source) until you are instructed to do so later in these procedures.

Is the system-attention LED on the control panel illuminated continuously?
- No: When the server is running, the system-attention LED is normally in the off state. However, you can manually put the LED in a blinking state to help identify the system unit. These states are normal and do not indicate a problem. To continue diagnosing a potential problem with your server, go to additional diagnostic procedures.
- Yes: A continuously illuminated system-attention LED means that the server has stored some information about a problem, and that you need to take an action to correct the problem. Continue with the next step.
Using care to not remove the power cord from the system unit or from the power source, remove the side cover from the system unit. See Remove and replace model 7037-A50 or 7047-185 covers and doors. After the side cover is removed, continue with the next step.

Locate the light-path-diagnostic card which is located on the disk drive cage. Is there a fault-indicator LED that is continuously illuminated on the light-path-diagnostic card?

Figure 1. 7037-A50 or 7047-185 light-path-diagnostic card Diagram of the light-path-diagnostic card

Table 1. 7037-A50 or 7047-185 Fault-indicator LEDs
1	Power supply fault-indicator LED	7	Front fan fault-indicator LED
2	Voltage-regulator module fault-indicator LED	8	Battery fault-indicator LED
3	Disk-drive bay fan fault-indicator LED	9	PCI adapter fault-indicator LED
4	Optical-media bay fault-indicator LEDs	10	Thermal fault-indicator LED
5	Disk-drive bay fault-indicator LEDs	11	Rear fan fault-indicator LED
6	System backplane fault-indicator LED	12	Memory fault-indicator LED

No: To continue diagnosing a potential problem with your server, go to the additional diagnostic procedures in step 8.
Yes: A continuously illuminated LED on the light-path-diagnostic card means that the server has detected a problem in the area of the server that is represented by the LED's position on the card. Continue with the next step.

Do one of the following actions:
Note: If you need to remove the air baffle, do not unplug the power cord from the server or the power source.
- If the thermal fault-indicator LED on the light-path-diagnostic card is illuminated and there are no other fault indicator LEDs illuminated, ensure that the air vents in the front and rear of the server are not blocked and that the room temperature is within the operational specifications.
- If the light-path-diagnostic card indicates that you have a problem with a component that is located under the air baffle, such as a memory module or a voltage-regulator module, you must open the door on the air baffle to see the component-identification LEDs, see Remove and replace model 7037-A50 or a 7047-185 covers and doors, then continue with step 5.
- If the light-path-diagnostic card indicates that you have a problem with the battery, power supply, front fan, rear fan, disk-drive fan, disk drive in one of the bays, or a PCI adapter in a PCI slot, go to step 6.
After the air baffle door is opened, you can see components such as memory modules and voltage-regulator modules and the component-identification LEDs that identify which module has a problem. Using the position of the fault-indicator LED from the previous step and the flashing component-identification LED, determine which module in your server is failing. Continue with the next step.
Call your service organization and report that one of the parts from the following list has a problem:
- memory module (DIMM)
- voltage-regulator module
- system backplane
- battery, Non-Volitile Random Access Memory (NVRAM)
- power supply
- PCI adapter fan
- processor fan
- disk drive enclosure fan
- media device in one of the bays
- disk drive in one of the bays
- PCI adapter in a PCI slot
Note: The preceding parts are listed with their part numbers in the Part number catalog.
When you receive the replacement part, continue with the next step.
Replace the part that was indicated by the fault-indicator LED. For removal and replacement instructions for the part, see Installing features and replacing parts. This ends the procedure.

You are here because your server has a problem that cannot be identified by the system-attention LED and fault-indicator LEDs. From the following table, find a symptom that describes your server's symptoms and do the related action:

Option	Description
Symptom	Action
The control (operator) panel is not illuminated.	Go to the Memory and processor subsystem problem isolation procedure.
The control (operator) panel is blank except for a dot moving across the control panel display.	The dot moving across the control panel display indicates that the control panel is hung in a reset state. This symptom might occur if a control panel cable is incorrectly seated or plugged, if a control panel or cable is damaged, or if the wrong control panel is installed. Do the following: Check and reconnect the control panel and cable connections. Refer to Model 7037-A50 and 7047-185 control panel If reconnecting the control panel cable did not fix the problem, replace the control panel. For control panel removal and installation instructions go to Model 7037-A50 and 7047-185 control panel. Note: After checking the above mentioned cables, plugs and control panel, and the symptom still has not been cleared, suspect the system backplane. Call your next level of support.
There is a five, six, or eight character code displayed in the control (operator) panel.	Go to Using system reference codes.
The server appears to have stopped. There is a three or four character code continuously displayed in the control (operator) panel for an unusually long period.	Go to Using progress codes.
*There is a code with a hyphen (xxx-xxxx) displayed in the control (operator) panel.*	You have a Service Request Number (SRN); go to AIX problem analysis.
The server appears to be looping. There is a series of three or four character codes repeatedly displaying in the control (operator) panel.	Go to Problems with loading and starting the operating system.
The AIX operating system is available but you still have a problem.	You have a problem with an AIX server; go to AIX problem analysis.
The Linux operating system is available but you still have a problem.	You have a problem with a Linux server; go to Linux problem analysis.
The replacement part does not fix the problem.	After the part is replaced, complete the system verification procedure. If your server shows the same symptoms, call your next level of support. If your server shows different symptoms than before, diagnose the new symptoms. Return to the first step of this procedure.
Find a System Reference Code (SRC) using the Operator Panel functions	Go to Control panel functions.
Find an SRC using the System Management Services	Go to Managing the system management services.
AIX diagnostics	Go to Working with AIX diagnostics.
Linux tool kit Web page	Go to Obtaining service and productivity tools for Linux.
Firmware update procedures	Go to Getting server firmware fixes for a 7037-A50 or 7047-185 system.